Abstract
Although anticoagulation is almost universally accepted as the preferred treatment for pulmonary embolism and for venous thrombosis in the lower extremities, its use is for the most part based on anecdotes and retrospective clinical surveys; data from properly controlled prospective studies are very few. By today's standards, the most frequently cited trials of anticoagulant drugs in pulmonary embolism1 , 2 are deficient in design, and the evidence supporting anticoagulant treatment for deep-vein thrombosis is equally unsatisfactory. When modern diagnostic methods have been employed, the purpose has usually been to compare anticoagulation for venous thrombosis with treatment by thrombolytic agents rather than with . . .